Research objectives and content
The subject of my training project will be the realisation of the most promising types of light sources and detectors for use in short-range optical interconnects. My prime criterion for the choice of the devices is their industrial relevance and applicability for inter-VLSI-chip interconnects. As explained at length in the Detailed Proposal, I am convinced that high-efficient LEDs are the most promising sources in this respect. In the limited time frame of the project, I will focus on one particular LED, the Non-Resonant Cavity (NRC)LED. My first contribution will be to decrease its present complexity and cost by introducing a Bragg reflector containing low-index oxidized AlAs layers, a material pioneered for VCSEL development. I will realise large NRC-LED arrays to demonstrate the success of this approach. My second contribution will be to study and improve the speed limits of NRC-LEDs to meet the demands of the envisaged application. Concerning detectors, I will concentrate on GaAs-based devices because they can be made on the same substrate as the emitters, which will ultimately simplify the integration with VLSI. My goal is to achieve GaAs-based detectors at 850 nm showing 70% efficiency at 300 MHz in array sizes of 256 elements. To achieve this, I will, for reasons of work efficiency, also make use of the Non-Resonant-Cavity.
Training content (objective, benefit, expected impact)
The proposed training project teaches how to combine profound scientific interest with high industrial applicability. The training center IMEC offers, in addition to this, a broader training in other fields. In combination, this unique training project will bring about the following benefits: - I will learn about optoelectronic devices on a scientific level, i.e. concerning growth with molecular beam epitaxy, processing and equipment, modelling, characterization, etc.... and in this way broaden my background in the field. - I will learn how to chose research subjects and strive for solutions which end in a product offering industrial applicability, i.e. reliability, cost-effectiveness, mass-production, packaging, thermal budgeting. It also includes remaining aware of competitor solutions (In the case of my project: The steady improvement of electrical interconnects). - in IMEC, I will be surrounded by a stimulating environment for broadening my views on current research topics in semiconductor industry. This includes Silicon VLSI design and processing, materials, packaging, analysis, etc...
The market of semiconductor LEDs is large, with a yearly production of more than 20 billion diodes for visible light-emitting diodes only. Cheap efficient LEDs are of great importance for such companies as Hewlett Packard and EGG, which have close contacts with the host institute IMEC. The problem of short-range optical interconnects is also of strategic interest for many industries. The recent MEL-ARI projects on the subject, e.g. contain European industrial partners such as Thomson-CSF, NMRC, GMMT, Framatome. IMEC being the prime contractor of OIIC, one of these projects, useful results can by this channel immediately be transferred to industry.